The A2780 ovarian cancer cell line was maintained in Roswell Park Memorial Institute (RPMI) Medium-1640 supplemented with: 10% Fetal Bovine Serum (FBS, Hyclone), 100 U/ml Penicillin, and 10 μg/ml Streptomycin. Cells were cultured in a humidified atmosphere of 5% CO₂ at 37°C, and the medium was changed every 48 h as described previously (Kakar et al., 2017). The A2780 ovarian cancer cell line is of human origin, necessitating the utilization of immune-deficient mice.

As described previously (Kelm et al., 2020; Straughn and Kakar, 2019), six-week-old female NOD.Cg-Prkdc^scid Il2rg^{TM 1Wjl}/SzJ (NSG, Jackson Lab Strain #005557) immunodeficient mice were initially randomly assigned to a tumor-free or tumor-bearing group (30 mice/group). Tumor-bearing groups received an intraperitoneal (i.p.) injection of 8.0 × 10⁵ low passage A2780 cells suspended in 100 μl sterile PBS. Control group received i.p. injection of 100 μl sterile PBS alone. After an initial refractory period of 8 days, mice in both the tumor-free and tumor-bearing groups were stratified into a group that would receive vehicle injections (10% dimethyl sulfoxide, 90% glycerol trioctanoate) or one of two concentrations of WFA (2 or 4 mg/kg) via i.p. injection (10 mice/group). Injections were performed once every 3 days until the culmination of the study. Post-euthanization, several tissues were collected, weighed, snap frozen in liquid nitrogen, and then stored at −80°C for further analysis. While alive, the mice were housed in a 12-h light–dark cycle and given water and food ad libitum. The Institutional Animal Care and Use Committee (IACUC, protocol # 15405) and Institutional Biosafety Committee (IBC, protocol # 18-208) of the University of Louisville approved all experimental protocols in mice in advance. None of the data in these experiments include data/samples from our initial publication discussing cachexia in the context of ovarian cancer (Straughn and Kakar, 2019).

Protocol used for the measurement of grip strength was essentially similar to as described previously (Gallot et al., 2018; Straughn and Kakar, 2019). Before assessment, mice were weighed on a commercially available digital scale. Forepaw and total grip strength of mice were measured using a digital grip strength meter (Columbus Instruments, Columbus, OH, United States) and then normalized by total body weight. Before the beginning of the test, the mice were acclimatized for five minutes. The mouse was allowed to grasp the total paw pull-bar assembly, and in a separate experiment the forepaw pull-bar assembly. The mouse was then gently drawn with constant force in a straight line away from the device until the mouse could no longer grasp the bar. Force at time of release was recorded as the peak tension. Each mouse was tested five times with a delay of 20–40 s between each testing. The mean peak tension was calculated from the recordings normalized by total body weight at time of recording. Forelimb and total limb grip strength were assessed on a weekly basis to balance elucidating functional muscle changes in response to tumor burden and WFA treatment, while simultaneously avoiding habituation to the grip strength analyses.

Isolation of total RNA from the gastrocnemius (GA) was performed using an RNeasy Fibrous Tissue Mini Kit (Qiagen Catalog # 74704) according to the manufacturer’s instructions. Skeletal muscle RNA was treated with an RNase-free DNase kit (RNase-Free DNase Set, Qiagen Catalog # 79254) in column according to the manufacturer’s instructions. First strand cDNA was synthesized using 1 μg of purified RNA and a commercially available kit (iScript^TM cDNA synthesis, Bio-Rad Catalog # 170-8891). Quantification of mRNA expression was performed similar to as previously described (Hindi and Kumar, 2016; Fong et al., 2012) using the SYBR Green dye method on a 7300 Real-Time PCR system (Applied Biosystems) using gene-specific primers, detailed in Supplementary Table 1.

Select muscles of the lower limb of the mice were isolated, flash frozen in liquid nitrogen, mounted in optimal cutting temperature (OCT) embedding medium, and then sectioned using a microtome cryostat. To assess tissue morphology, 10 μm thick transverse sections were cut from the mid-belly of the tibialis anterior (TA) and then subjected to Hematoxylin and Eosin (H&E) staining. Images of H&E-stained TA muscle sections were quantified using Fiji software (National Institute of Health software) to measure myofiber CSA. Skeletal myofiber CSA was calculated by analyzing ∼500–700 myofibers per muscle as previously described (Kakar et al., 2012).

To detect changes in the number and myogenic status of satellite cells, 8 μm thick transverse sections were cut from the mid-belly of the TA. Slides were fixed in freshly prepared 3.7% Formalin solution for 5 min. Slides were subsequently washed in 0.3 M Glycine in PBS for 10 min. Slides were then briefly washed in 0.1% Triton in PBS. Slides were then incubated for 10 min in 0.1% Triton in PBS. Slides were then subjected to an antigen retrieval step. Briefly, slides were incubated in Coplin staining jars containing 0.01 M citrate buffer and placed inside of a pressure cooker. Slides were incubated on “High Pressure” for 10 min. Slides were cooled and then washed in PBS. Slides were then blocked in a PBS solution containing 3% bovine serum albumin, 2% horse serum, and Mouse-on-Mouse blocking reagent (1 drop per 1.25 ml of blocking solution) for 60 min in a humidified chamber. Slides were then incubated overnight at 4°C in blocking solution containing primary antibodies against Pax7, MyoD, and Laminin proteins. The next day, slides were washed three times with PBS. Following, slides were incubated for 60 min in blocking solution containing appropriate secondary antibodies conjugated to fluorophores with minimally overlapping spectral emissions. Slides were subsequently washed three times using PBS for 5 min. Slides were then incubated for 5 min in DAPI resuspended in PBS (1:5,000). Slides were subsequently washed three times using PBS for 5 min. Refer to Supplementary Table 2 for a complete list of antibodies utilized. Four images from distinct areas of the muscle were captured for each animal. Imaging and quantification were performed by two separate people to allow for blinding of group identity through the use of an alphanumeric identifier. Group identities were unmasked after all data was quantified and submitted to the principal investigator, such that group-wise comparisons could be performed. For the analysis, the number of Laminin + myofibers was first counted, such that the data could eventually be normalized between images to account for differences in the size/area imaged. Next, distinct puncta in the Texas Red channel (corresponding to the secondary antibody employed for Pax7) were annotated on the image using Fiji software. These marks were then compared with the DAPI channel. When the mark for the Texas Red channel did not overlap with nuclear staining on the DAPI channel, the prospective satellite cell was ruled as a false positive and the identifying mark on the image was deleted. After deletion of false positives from the cell count feature, the number of satellite cells for the respective image was recorded. The frequency of Pax7+ cells per laminin + myofiber was then calculated for each image and then averaged. Next, the marks from the Texas Red channel were compared against the YFP channel (corresponding to the secondary antibody employed for MyoD). Two additional cell counts were initialized such that we could count cells that also exhibited YFP overlap (i.e., satellite cells that were Pax7+/MyoD+) vs. cells that did not have a YFP spectral emission (i.e., satellite cells that were Pax7+/MyoD-). The proportion of differentiating satellite cells and the proportion of satellite cells that were proliferating/quiescent was calculated for each of the four images and then averaged. This process was repeated for all 60 mice (totaling 240 RGBY images).

Slides were mounted using Eukitt Quick-hardening mounting medium (Sigma-Aldrich) and visualized at −0.4°C on a Nikon TiE 3000 inverted microscope (Nikon) equipped with a digital camera (DS-U2/L2-Ri1 digital microscope camera (Nikon) for light microscopy or DXM-1200C coded digital camera (Nikon) for fluorescent microscopy), and Nikon NIS Elements AR software (Nikon). Exposure times were consistent for each staining type. Image levels were equally adjusted using Adobe Photoshop CS6 software (Adobe) to remove non-specific background staining. Margins of cropped images are indicated by a dashed red border or a solid black line.

Quadriceps femoris (QF) samples were homogenized in chilled RIPA buffer (Sigma) supplemented with a Complete Mini Protease Inhibitor tablet (Roche Molecular Biochemicals, Indianapolis, IN, United States). Tissue lysates were centrifuged at 10,000 RPM and the supernatants were collected. Protein concentration for each sample was determined using the Bradford reagent method (Bio-Rad), according to manufacturer’s instructions. Protein lysates (50 μg) were separated on 10% SDS-PAGE gels at 100 V for 2 h. The proteins were transferred to nitrocellulose membranes at 100 V for 90 min. The membranes were blocked with 5% non-fat milk in Tris Buffered Saline supplemented with Tween-20 (0.05%; TBS-T) for 30–60 min. The membranes were then washed three times with TBS-T for 5 min. The membranes were then incubated with primary antibody at 4°C overnight. The membranes were washed three times with TBS-T for 5 min each followed by incubation in TBS-T containing horseradish peroxidase conjugated secondary antibody (1:3,000 dilution) for 1 h. The membranes were rinsed three times with TBS-T for 5 min each. Visualization of immunoreactive bands was enhanced using chemiluminescence reagents (Sigma). The membranes were stripped off using Restore^TM Western Blot Stripping Buffer (Thermo Scientific Catalog # 21059) for 15 min, blocked, and re-probed with horseradish peroxidase-conjugated β-Actin or GAPDH as a control to normalize loading variation. Refer to Supplementary Table 2 for a complete list of antibodies utilized.

For the sake of transparency, the majority of the results were expressed as box-and-whisker plots with the box comprised of the first, second, and third quartiles, and the lower and upper whiskers corresponding to the minimum and maximum values, respectively, to display the entire range of data. Individual data points are depicted as black circles. Cropping of images is indicated by a dashed red border or a solid black line around the field of view. Statistical analysis of the data was performed using an unpaired two-tailed t-test with Welch’s correction for simple two group comparisons, a one-way analysis of variance (ANOVA) followed by Tukey’s Honestly Significant Difference Test (HSDT) post hoc analysis for comparisons between 3 or more groups with one experimental factor, or a two-way ANOVA followed by Tukey’s multiple comparison test post hoc analysis for comparisons between 4 or more groups containing two experimental factors to determine statistically significant differences between groups with GraphPad Prism 8.3.0 software for Mac (La Jolla, CA, United States). ANOVA summaries and the results of the multiple comparison tests are presented in Supplementary Material. Welch-corrected or Tukey-corrected p-value of <0.05 was considered statistically significant, unless otherwise specified.

All procedures involving the usage of mice were carried out in strict accordance to the standards of the National Institute of Health guide for the care and use of laboratory animals. The Institutional Animal Care and Use Committee (IACUC, protocol # 15405) and Institutional Biosafety Committee (IBC, protocol # 18-208) of the University of Louisville approved all experimental protocols in mice in advance. No human data or tissue was used in this study.

Recent work from our lab has demonstrated that the A2780 ovarian cancer cell line is capable of inducing a skeletal muscle cachectic phenotype and that WFA attenuates these changes (Straughn and Kakar, 2019). In the present study, we attempted to elucidate whether WFA has an effect on functional muscle strength and the size of muscle in tumor-free mice, and to corroborate our prior results in tumor-bearing mice using the same lower dosage of WFA (2 mg/kg), but a different upper dosage of WFA (4 mg/kg) due to the deleterious effects previously observed with 6 mg/kg of WFA. Additionally, due to the previously reported tumor-associated mortality, the amount of xenografted cells was reduced from 1 × 10⁶ to 8 × 10⁵. A parallel report to this study from our lab examining gross body changes, tumor burden, and the effect on cardiac muscle in response to the xenografting of the A2780 ovarian cancer cell line and WFA treatment has recently been discussed in the work by Kelm et al. (2020). Information pertinent to this manuscript can be found in Figure 1, Supplementary Figure 1 and Supplementary Table 1 of the aforementioned manuscript. Before the mice received i.p. xenografts of the A2780 ovarian cancer cells, we performed forelimb and total grip strength analyses to ensure no significant differences within the groups stratified to become tumor-free or tumor-bearing existed at the start of the study (Supplementary Figures 1A,B). As expected, no significant differences were found in the baseline forelimb (tumor-free: 0.051 ± 0.004 N/g; “tumor-bearing”: 0.052 ± 0.004 N/g) or total grip strength (tumor-free: 0.094 ± 0.006 N/g; “tumor-bearing”: 0.092 ± 0.007 N/g) in the groups that were initially stratified to become tumor-free or tumor-bearing animals, as determined by an unpaired two-tailed Welch-corrected t-test [forelimb: t(57.63) = 0.74, p = 0.47; total limb: t(54.92) = 1.27, p = 0.21] (Supplementary Figures 1A,B). Subsequently, mice were examined for changes in grip strength once per week until the culmination of the study. Based upon the mortality rate observed in our previous study (Straughn and Kakar, 2019), the endpoint of this study was set at week 4 post-xenografting of the ovarian cancer cells. One-week post-xenografting, before WFA treatment was initiated, we observed a significant reduction in the forelimb (tumor-free: 0.056 ± 0.003 N/g; tumor-bearing: 0.049 ± 0.003 N/g) and total grip strengths (tumor-free: 0.096 ± 0.004 N/g; tumor-bearing: 0.088 ± 0.006 N/g) of the tumor-bearing mice compared to the tumor-free control group [forelimb: t(56.62) = 8.02; total limb: t(52.64) = 6.03; p < 0.0001 for both comparisons], suggesting a rapid onset of muscle decline (Supplementary Figures 1C,D).

At the terminal week of the study, we observed that treatment with WFA at 2 and 4 mg/kg significantly improved the forelimb grip strength of tumor-free mice (Figure 1A), and that the higher dosage of WFA had significantly improved the total grip strength of the tumor-free mice (Figure 1B) compared to the tumor-free vehicle-treated group as determined by a two-way ANOVA followed by Tukey’s multiple comparison test post hoc analysis (forelimb – WFA 2 mg/kg: p = 0.02; WFA 4 mg/kg: p < 0.0001; total limb – WFA 2 mg/kg: p = 0.30; WFA 4 mg/kg: p = 0.0005), suggesting that WFA treatment improves basal grip strength. Similar to our recently reported study (Straughn and Kakar, 2019), the forelimb and total limb grip strength of the tumor-bearing vehicle-treated group was significantly reduced compared to the tumor-free groups (p < 0.0001 for all comparisons) (Figures 1A,B). The tumor-bearing WFA-treated groups displayed forelimb and total grip strengths that were significantly increased compared to the tumor-bearing vehicle-treated (p < 0.0001 for all comparisons), but not significantly different than the tumor-free vehicle-treated group, corroborating our prior report that WFA ameliorates the weakening effects on skeletal muscle in our xenograft model of ovarian cancer-induced cachexia (Figures 1A,B). Grip strength analyses for week 2 and week 3 post-xenografting are presented in Supplementary Figures 1E–H. These intermediate timepoints demonstrate similar trends and p-values as the terminal week of the study.

Post-mortem, we collected select muscles of the lower extremities (TA, GA, and QF) and weighed them to detect changes in muscle mass (Figure 1C). Muscle weights were normalized by the initial body weight (IBW) to account for differences in the size of the mouse/muscle at baseline, while excluding the confounding effect of tumor burden on body weight. A reduction in the normalized weight of the TA, GA, and QF muscles was observed in the tumor-bearing vehicle-treated group compared to the tumor-free vehicle-treated group (p < 0.05 for all comparisons) (Figure 1C). Within the tumor-free groups, the WFA 4 mg/kg group displayed a significant increase in normalized weight of the TA, GA, and QF muscles compared to the vehicle-treated group (p < 0.05 for all comparisons) (Figure 1C). Within the tumor-bearing groups, treatment with both concentrations of WFA led to a statistically significant increase in normalized wet weight of the TA and GA muscles compared to the vehicle-treated group (Figure 1C). Curiously, these results are in contrast with those from our prior publication, where we did not observe significant differences in muscle mass (Straughn and Kakar, 2019). Experimental differences between the two studies, such as the initial cancer cell xenograft volume and duration of treatment could account for the seemingly discrepant results. It is also possible that the relatively small sample size in the first study did not allow us to observe small differences in muscle mass.

In an attempt to corroborate the changes in muscle strength and weight in response to WFA treatment and the ovarian cancer cells, changes in myofibrillar size in the TA muscle were examined via H&E staining (Figure 2A). Similar to the changes in grip strength observed, both concentrations of WFA led to a significant increase in the average myofibrillar CSA in the TA muscle of the tumor-free WFA-treated groups (WFA 2 mg/kg: 1,916.39 ± 85.72 μm²; WFA 4 mg/kg: 1,904.98 ± 80.50 μm²) compared to the tumor-free vehicle-treated group (1,652.48 ± 46.13 μm²) (p < 0.0001 for both comparisons) (Figure 2B). The average CSA of the tumor-bearing vehicle-treated group (1,097.28 ± 74.60 μm²) was significantly decreased compared to all tumor-free groups (p < 0.0001 for all comparisons) (Figure 2B). Within the tumor-bearing groups, both dosages of WFA led to a complete rescue in myofibrillar CSA (WFA 2 mg/kg: 1,629.72 ± 94.96 μm²; WFA 4 mg/kg: 1,748.13 ± 68.90 μm²) compared to the tumor-bearing vehicle-treated group (p < 0.0001 for both comparisons) (Figure 2B). The average myofibrillar CSA of the tumor-bearing WFA-treated groups was not significantly different from that of the tumor-free vehicle-treated group (WFA 2 mg/kg: p = 0.99; WFA 4 mg/kg: p = 0.07) (Figure 2B). To confirm accurate measurement of the CSA, the minimal Feret’s diameter was measured in conjunction with the CSA (Figure 2C). Nearly identical trends and levels of significance in the minimal Feret’s diameter were observed compared to the CSA, indicating the validity of our histological assessment (Figure 2C). While we observed changes in myofiber size, muscle mass, and grip strength, it cannot be stated definitively that these are linked. Indeed, whether or not a correlation between muscle size and muscle functions exists is currently unresolved and under intense debate within the current literature (Loenneke et al., 2019; Taber et al., 2019).

Dysregulation of NF-κB signaling has been shown to spuriously activate satellite cells and lead to an impairment of skeletal muscle repair (He et al., 2013; Straughn et al., 2019). Previous work from our lab demonstrated that WFA treatment results in a reduction in activation/nuclear translocation of the canonical NF-κB signaling pathway component p65 within the xenografted tumor (Straughn and Kakar, 2019). However, we did not assess changes in NF-κB signaling within skeletal muscle in our prior study. Thus, we performed western blotting against phospho- and total levels of p65 in protein extracts from the QF muscle (Supplementary Figures 2A,B). We observed a significant increase in the ratio of phosphorylated p65 to total p65 protein in the tumor-bearing vehicle-treated group (Supplementary Figure 2B). This increase in NF-κB activation was ameliorated upon treatment with WFA group (Supplementary Figure 2B). Next, we assessed for changes in the IKK complex and RelA at a transcriptional level in the GA muscle (Supplementary Figure 2C). The relative transcript levels of RelA and IKKβ were significantly increased in the tumor-bearing vehicle-treated group compared to the tumor-free vehicle-treated group (Supplementary Figure 2C). As expected, WFA treatment in both tumor-free and tumor-bearing animals led to a significant reduction in relative transcript levels of RelA and IKKβ compared to their vehicle-treated controls (Supplementary Figure 2C). Due to WFA’s known role in inhibiting canonical NF-κB signaling (Heyninck et al., 2014; Kaileh et al., 2007), and the effect ovarian cancer has on this signaling pathway (Chen et al., 2008), we investigated if WFA or our xenograft model induced the activation of satellite cells.

The TA muscle was subjected to immunofluorescent immunohistochemical (IHC) staining for Pax7 (a marker of satellite cells), MyoD (a marker of myogenic differentiation), and Laminin (a marker of the basal lamina) (Figure 3A). As accurate IHC detection of satellite cells can be challenging, individual color channels were compared against negative control slides to optimize visualization of satellite cells, minimize non-specific background staining, and rule out false positives (Pax7⁺/DAPI^– cells) (Feng et al., 2018). Initially, the number of satellite cells per field (normalized by the number of Laminin⁺ myofibers) was quantified to assess whether our model of ovarian cancer or WFA treatment affected the satellite cell population at large (Figure 3B). Within the tumor-free groups, WFA treatment significantly increased the normalized amount of Pax7⁺ cells (WFA 2 mg/kg: 0.28 ± 0.02; WFA 4 mg/kg: 0.42 ± 0.01) compared to the tumor-free vehicle-treated group (0.06 ± 0.01) (p < 0.0001 for both comparisons) (Figure 3B). There was also a significant increase in the amount of Pax7⁺ cells in the tumor-bearing vehicle-treated group (0.34 ± 0.01) compared to the tumor-free vehicle-treated group (p < 0.0001) (Figure 3B). Interestingly, within the tumor-bearing groups, the WFA 2 mg/kg group displayed a significant reduction in Pax7⁺ cells (0.26 ± 0.01) compared to the vehicle-treated and WFA 4 mg/kg groups (0.31 ± 0.01) (p < 0.0001 for both comparisons) (Figure 3B). No definitive reason for this experimental observation has currently been elucidated. A small, but statistically significant decrease in the tumor-bearing WFA 4 mg/kg group was observed compared to the tumor-bearing vehicle-treated group (p = 0.0004) (Figure 2B). The tumor-free WFA 4 mg/kg group displayed the highest proportion of Pax7⁺ cells, suggesting that WFA is a potent activator of satellite cells.

After assessing changes in the gross number of satellite cells, we investigated their myogenic status to determine if the satellite cells were functionally activated. Satellite cells that are Pax7⁺/MyoD^– are self-renewing or returning to quiescence, whereas satellite cells that are Pax7⁺/MyoD⁺ have committed to the myogenic lineage and will facilitate in muscle repair (Kuang and Rudnicki, 2008; Kuang et al., 2007). Similar to published reports (Didier et al., 2012), the vast majority of nuclei within and outside of myofibers were found to be Pax7^–/MyoD⁺, and as such were not considered myogenic progenitors. It was not feasible to enumerate the population of Pax7^–/MyoD⁺ cells in response to the xenografted cancer or WFA treatment due to experimental limitations. However, distinct populations of Pax7⁺/MyoD^– and Pax7⁺/MyoD⁺ cells were present (Figures 3C,D). Consistent with published reports (Yoshida et al., 1998), the majority of the satellite cells in the tumor-free vehicle-treated group were Pax7⁺/MyoD^– (96.29 ± 0.69%) and a small population was found to be Pax7⁺/MyoD⁺ (3.71 ± 0.69%) (Figures 3C,D). The proportion of proliferating satellite cells in the tumor-bearing vehicle-treated group (80.10 ± 1.88%) was significantly decreased, and the proportion of differentiating satellite cells (19.90 ± 1.88%) was significantly increased compared to the tumor-free vehicle-treated group (Figures 3C,D). In the tumor-free groups, there was a significant decrease in the proportion of proliferating satellite cells (WFA 2 mg/kg: 39.27 ± 2.28%; WFA 4 mg/kg: 38.63 ± 1.49%) and a significant increase in differentiating satellite cells (WFA 2 mg/kg: 60.73 ± 2.28%; WFA 4 mg/kg: 61.37 ± 1.49%) in the WFA-treated groups compared to the vehicle-treated group (p < 0.0001 for all comparisons) (Figures 3C,D). In the tumor-bearing groups, there was a significant decrease in the proportion of proliferating satellite cells (WFA 2 mg/kg: 51.15 ± 1.63%; WFA 4 mg/kg: 47.21 ± 1.99%) and a significant increase in differentiating satellite cells (WFA 2 mg/kg: 48.85 ± 1.63%; WFA 4 mg/kg: 52.79 ± 1.99%) in the WFA-treated groups compared to the tumor-free and tumor-bearing vehicle-treated groups (p < 0.0001 for all comparisons) (Figures 3C,D).

To corroborate our IHC data, we performed qPCR on GA muscle extracts to assess relative transcript levels of Pax7 and Myod1 (Figure 3E). Similar to the IHC data, the tumor-bearing vehicle-treated group displayed a significant increase in relative transcript levels of Pax7 and a significant decrease in Myod1 compared to the tumor-free vehicle-treated group (Figure 3E). The tumor-free and tumor-bearing WFA-treated groups displayed significant increases in Pax7 compared to the tumor-free vehicle-treated group and a significant increase in Myod1 compared to both vehicle-treated groups (Figure 3E). While both WFA treatment and xenografting of ovarian cancer into mice led to an increase in satellite cell proliferation, WFA appears to signal for the differentiation of satellite cells through a mechanism not presently known.

It has been demonstrated in the Lewis lung carcinoma and Apc^Min/+ models of cancer-induced cachexia that several markers of the UPR are upregulated compared to tumor-free hosts as a byproduct of increased endoplasmic reticulum (ER) stress (Bohnert et al., 2016). Due to the UPR’s role in regulating muscle mass and common perturbations associated with cancer-induced cachexia, we investigated if components of the UPR are augmented in response to WFA treatment or xenografting of the A2780 ovarian cancer cell line. Overall, both WFA and the xenografted ovarian cancer cell line led to activation of various components of the UPR in skeletal muscle, with the A2780 ovarian cancer cell line leading to a higher degree of activation than WFA treatment (Figure 4).

The relative transcript levels of Eif2ak3 [encoding the protein PERK (Hayes et al., 1999)], Atf4 [encoding the protein ATF4 (Karpinski et al., 1992)], Ddit3 [encoding the protein CHOP (Marola et al., 2019)], Ppp1r15a [encoding the protein GADD34 (Clavarino et al., 2012)], Hspa5 [encoding the protein GRP78 (Elfiky et al., 2020)], Hsp90B1 [encoding the protein GRP94 (Chen et al., 2005)], and Tnfrsf10B [encoding death receptor 5 (DR5; Dittmann et al., 2020)] were significantly increased in the GA muscle of the tumor-bearing vehicle-treated group compared to the tumor-free vehicle-treated group (p < 0.0001 for all comparisons) (Figure 4). Curiously, relative transcript levels of Ern1 [encoding inositol requiring enzyme 1α (IRE1α; Tirasophon et al., 1998)] were significantly reduced in the tumor-bearing vehicle-treated group compared to the tumor-free vehicle-treated group (p = 0.001) (Figure 4).

In the tumor-free mice, we observed: (1) transcriptional upregulation of Eif2ak3, Ddit3, Ppp1r15a, and Ern1, (2) downregulation of Hspa5 and Hsp90B1, and (3) no change in transcript levels of Atf4 and Tnfrsf10B in the WFA-treated groups compared to the vehicle-treated group (p < 0.001 or 0.0001 for most comparisons) (Figure 4). In the tumor-bearing mice, the WFA treatments produced a mixture of (dis)similar results. In the tumor-bearing WFA 2 mg/kg group, we observed: (1) an upregulation of relative transcript levels of Ddit3 (p < 0.01), (2) a reduction in Eif2ak3, Atf4, Ppp1r15a, Hspa5, Hsp90B1, and Tnfrsf10B (p < 0.0001 for all comparisons), and (3) no significant difference in relative transcript levels of Ern1 (p = 0.9016) compared to the tumor-bearing vehicle-treated group (Figure 4). In the tumor-bearing WFA 4 mg/kg group, we observed: (1) a significant increase in relative transcript levels of Ddit3 and Ern1 (p < 0.0001 for both comparisons) and (2) a decrease in relative transcript levels of Eif2ak3, Atf4, Ppp1r15a, Hspa5, Hsp90B1, and Tnfrsf10B (Eif2ak3: p = 0.0085; p < 0.001 for all other comparisons) compared to the tumor-bearing vehicle-treated group (Figure 4). The expression of Atf6 was not detectable in a sufficient number of samples within the tumor-free or tumor-bearing groups, therefore no conclusions were drawn for this branch of the UPR (Data not shown).

Subsequently, activation of the PERK and IRE1α arms of the UPR were assessed by Western blotting for both the phosphorylated and total protein levels of PERK and IRE1α in an attempt to corroborate the changes observed at the transcriptional level (Figures 5A,B). The ratio of phosphorylated PERK to the total levels of PERK was significantly increased in the tumor-bearing vehicle-treated group compared to the tumor-free groups, suggesting that the xenografted ovarian cancer results in activation of PERK signaling (Figure 5B). This increase in PERK activation was reversed upon treatment with WFA (4 mg/kg) and did not differ from the levels of PERK activation in the tumor-free vehicle-treated group (Figure 5B). To further assess PERK activation, western blotting against the protein CHOP was performed; however, markedly variable results were obtained in all of the groups assessed (Data not shown). The ratio of phosphorylated IRE1α to the total levels of IRE1α was significantly increased in both the tumor-free and tumor-bearing WFA 4 mg/kg groups compared to their vehicle-treated counterparts, suggesting that WFA leads to the activation of the IRE1α arm of the UPR (Figure 5B). Limitations of this experiment include the necessity to stratify macromolecular extraction to different muscles of the lower extremity due to the atrophying effects of the cancer on skeletal muscle.

In response to ER stress, IRE1α catalyzes the splicing of XBP-1 mRNA (Bohnert et al., 2018). To further assess activation of the IRE1α arm of the UPR, we evaluated the splicing of XBP-1 by performing semi-quantitative RT-PCR using a set of primers that detects both the unspliced (uXBP-1) and spliced XBP-1 (sXBP-1) mRNA variants (Figure 5C). We also assessed total levels of XBP-1 using a primer sequence that detects a conserved region shared between the spliced and unspliced variants of XBP-1, as well as the unrelated gene β-Actin to use as a control. Representative images for these experiments are presented in Figure 5C. The ratio of sXBP-1 to uXBP-1 was significantly increased in all of the WFA-treated groups, and a trend toward increase in the tumor-bearing vehicle-treated group (p = 0.06) was observed compared to the tumor-free vehicle-treated group (Figure 5D). After normalization by β-Actin, no significant differences in the total amount of XBP-1 was observed between any of the groups (Figure 5D). Based upon the qPCR data and the western blotting results, it appears that WFA and our model of ovarian cancer are antagonistically acting upon the UPR to induce an adaptive or maladaptive response, respectively.

Downstream of the UPR, the UPS is one of the two major proteolytic systems that degrades muscle proteins in oncological settings (Sandri, 2016). As an initial experiment to assess for changes in the UPS, we performed immunoblotting against ubiquitin to detect pan-levels of ubiquitination (Figure 6A). As expected, there was a significant increase in the amount of proteins in the QF muscle conjugated to ubiquitin in the tumor-bearing vehicle-treated compared to the tumor-free vehicle-treated group (Figure 6B). WFA treatment (4 mg/kg) significantly reduced the levels of ubiquitinated proteins in the tumor-free and tumor-bearing groups compared to their respective vehicle-treated control group.

We next investigated relative transcript levels of muscle-specific E3 ubiquitin ligases, as well as an E3 ubiquitin ligase that is not specific to muscle [Traf6, encoding the protein TRAF6 (Cao et al., 1996)] (Bodine and Baehr, 2014; Paul et al., 2012). Fbxo30 [encoding the protein MUSA1 (Hahn et al., 2020)], Fbxo32 [encoding the protein MAFbx (Hahn et al., 2020)], and Trim63 [encoding the protein MuRF1 (Hahn et al., 2020)] are muscle-specific E3 ubiquitin ligases that have been shown to be involved in protein degradation under catabolic settings (Bodine and Baehr, 2014; Foletta et al., 2011). Within the tumor-free groups, there was a significant reduction in the relative transcriptional expression of Traf6 and Fbxo30 in the WFA 2 mg/kg group compared to the vehicle-treated group (p < 0.05 for both comparisons) (Figures 6C,D). The WFA 4 mg/kg group exhibited a significant reduction in relative gene expression of Fbxo30, Fbxo32, Trim63, and Traf6 compared to the tumor-free vehicle-treated group (p < 0.0001, p = 0.04, p = 0.02, p < 0.0001, respectively) (Figures 6C,D). The tumor-bearing vehicle-treated group exhibited a significant upregulation in relative transcript levels of all markers of the UPS assessed compared to all tumor-free groups (p < 0.0001 for all comparisons) (Figures 6C,D). In the tumor-bearing animals, treatment with WFA led to a statistically significant dose-dependent transcriptional downregulation of Fbxo30, Fbxo32, Trim63, and Traf6 compared to the tumor-bearing vehicle-treated group (p < 0.0001 for all comparisons) (Figures 6C,D). The results of these experiments suggest that modulation of the UPS could be one of the signaling pathways responsible for WFA’s positive effect on skeletal muscle mass.

In addition to the UPS, cancer-induced cachexia has been shown to activate the ALS to facilitate an atrophying effect on skeletal muscle (Sandri, 2010). During autophagy, LC3B-I is converted to LC3B-II to allow interaction with autophagic vesicles (Tanida et al., 2004). p62 is a substrate of autophagy that recognizes and binds to ubiquitinated proteins (Lamark et al., 2017). As such, p62 protein levels decrease with the activation of autophagy and it becomes transcriptionally upregulated to replenish the depleting protein (Lamark et al., 2017). Beclin1 is another autophagy-related protein that is critical for the initiation of the autophagosome (Sun et al., 2009).

To examine the effect of WFA and our xenograft model on the ALS, we measured relative mRNA levels of common autophagy markers: Sqstm1 [encoding the protein p62 (Sánchez-Martín and Komatsu, 2018)], Map1lc3b [encoding the protein LC3B (Kang et al., 2019)], and Becn1 [encoding the protein Beclin1 (Zheng et al., 2020)] (Figure 7A). A significant reduction in relative gene expression of the aforementioned markers of the ALS was observed in the tumor-free WFA 4 mg/kg group compared to the tumor-free vehicle-treated group (p = 0.04, p = 0.013, p < 0.0001, respectively) (Figure 7A). In the tumor-bearing vehicle-treated group, we observed a significant increase in relative transcript levels of Sqstm1, Map1lc3b, and Becn1 compared to all tumor-free groups (p < 0.0001 for all comparisons) (Figure 7A). Similar to the results for the UPS, WFA treatment led to a significant dose-dependent reduction in relative mRNA expression of the select ALS markers compared to the tumor-bearing vehicle-treated group (p < 0.0001 for all comparisons) (Figure 7A).

Subsequently, we performed immunoblotting against LC3B, p62, and Beclin1 to determine whether WFA or ovarian cancer augmented signaling to the ALS at a protein level (Figures 7B,C). The ratio of LC3B-II to LC3B-I was significantly increased in the tumor-bearing groups compared to the tumor-free groups. However, WFA treatment did not affect LC3B lipidation under tumor-free or tumor-bearing conditions. Interestingly, protein levels of p62 were significantly increased in the tumor-free WFA 4 mg/kg compared to the tumor-free vehicle-treated group, suggesting a reduction in basal levels of autophagy (Figure 7C). A significant reduction in the levels of p62 in the tumor-bearing vehicle-treated and WFA 4 mg/kg groups was observed compared to the tumor-free vehicle-treated group, suggesting an active autophagy process was occurring (Figure 7C). Further, we observed a significant increase in Beclin1 only in the tumor-bearing vehicle-treated group compared to the tumor-free vehicle-treated group (Figure 7C). The tumor-bearing WFA 4 mg/kg group had a significant reduction in Beclin1 protein levels compared to the tumor-bearing vehicle-treated group (Figure 7C). Similar to the UPS, the results from these experiments suggest that ovarian cancer results in a hyper-catabolic state. However, WFA appears to only regulate the ALS at a transcriptional level.